SBIR-STTR Award

There exists today, a paucity of cost effective methods in bilingual education. An innovative approach is proposed which utilizes an intelligible, cost effective computer synthesized speech module possessing:-programmability in English, Spanish, German, and possibly other languages.-phonetic accuracy of computer pronounced words through a user maintained exception dictionary. -little or no maintenance requirements.The proposed approach will establish basic technology for providing a synthetic speech capability in computer software for bilingual education. NOVCOM Systems has already developed a state of the art text to speech (TTS) algorithm for English and proposes under this contract to address the feasibility of extending this technology to other languages. Today, Hispanics comprise the largest non-English speaking minority group in the United States. It is therefore statistically most prudent to consider the Spanish language for a bilingual education approach. NOVCOM will design a set of TTS translation rules for the Spanish language. For the purpose of concept demonstration, the English Spanish speech capability will be incorporated into a prototype tutorial program featuring computer synthesized English and Spanish speech.The anticipated results of this program are the development of an innovative computerized methodology for bilingual education. Phase 1 will concentrate on demonstrating concept feasibility and viability of an English Spanish speech module. Phase 2 will involve on-site testing of the Phase I prototype and extension of the methodology to additional languages. If Phases I and II are successful, sufficient groundwork will have been established for the introduction of a cost effective, marketable educational software package.

Under Part I of this initiative NOVCOM Systems, Inc.demonstrated feasability of a state of the art, bilingual, textto speech (TTS) methodology for the English and Spanishlanguages. This innovative development utilizes an intelligible,cost effective computer synthesized speech module possessing:- Convenient programmability in either English or Spanish.- Phonetic accuracy of computer pronounced words through a usermaintained exception dictionary.- Little or no maintenance requirements.Results of a synthesized word recognition test conducted at theUniversity of Minnesota yielded high levels of Spanish wordrecognition, indicating that the feasability of the Phase Iapproach has been demonstrated. During Phase II, field testingwith Hispanic and Latino and English students at all age andability levels, including handicapped students and adults, theEnglish, Spanish speech capability will be incorporated intoprototype tutorial courseware. This will enable computer lessonspresently on the terminal screen to be heard as well as seen. Amarketing strategy for the Phase II courseware is described.Phase II products have great potential for markedly increasinginstructional effectiveness.Anticipated Results and

Potential Commercial Applications:
Theanticipated results of this program are the development of aninnovative computerized methodology for bilingual education.Phase 2 will involve:- on-site testing of the Phase I prototype English, Spanishspeech module- extension of the methodology to the German language.- developing prototype instructional Spanish lessons.- testing the instructional effectiveness of Spanish lessons.The potential for success is a breakthrough in the education ofHispanic and Latino students such that they will learn basicskills and knowledge in a more cost effective and educationallyeffective manner than at present. The size of the market for thesimple computer hardware product and for software development isenormous-some 6,000,000 Hispanic and Latino speaking elementaryand secondary students in America alone are potentialbeneficiaries as clients, as well as adults, students in othercountries and the development of additional modules for otherlanguages.